Latest Contributions by DuarteCarreira

Dan, thanks for your reply, but I had already seen those links. What I'm hoping for is a more hands-on help. Yes, a howto kind of answer. The first link is pretty generic info, and the 2nd lists a bunch of config files which will help those who already have a deep understanding of the config mechanics... ... View more

Answer to your question: * Is it possible to use native PostGIS functions and geometry type in ArcMap without installing SDE stuff? It is possible to read, impossible to write/edit. Elaborating: You can create data in PostGIS using non-esri software like QGIS, and read it in ArcMap. You can add it to the map using menu File->Add Data->Add Query Layer. You can filter, symbolize, label, etc. You cannot edit. Option 2: you can only create PostGIS data using ArcMap if you install sde, then you can read it in QGIS. You cannot edit in QGIS. (you can but that's a bit more complicated because of ObjectID maintenance/creation) Option 3: you can create PostGIS data using ArcMap (with sde), publish a T-WFS service with ArcGIS Server, and edit the data in QGIS by using the t-WFS service. Option 4: buy an add-in for ArcMap that edits "pure" PostGIS data. ... View more

Better answers here: Map service connection to PostgreSQL DB shown as "Idle in Transaction" - Normal? It seems 10.4 changed to IDLE connections instead. But pooling is not a supported scenario since it could in some scenarios create unexpected results - worse case data corruption. Still, read-only connections could be a good candidate for pooling, even though it's not supported. ... View more

Philip thanks again for your insights. I'm glad to hear that esri is looking into this. I had thought about things that could be broken with connection pooling. I tested 2 PCs connecting with the same user, and making large selections - everything worked fine AFAIK. But the doubt was in my head now... One thing that would be a very big help and may be easy to support is read-only connections. Some 80% of our connections to the gdb are read-only. At least 40 connections are read-only arcgis server services. These only seat there eating away our server's memory. So just reducing these would make a big difference. Same can be said of desktop users. Only a few have editing privileges. Big majority only reads data, although ofcourse writing stuff like big selection sets. What are your thoughts on this? I'm not looking for a certified answer from esri, just for a good discussion on this issue. I understand things move at their own pace. Thanks again. ... View more

Philip Heede​ thanks for the heads up. Can you elaborate a bit more? I have made a simple test with desktop 10.4 and pgbouncer, opening an mxd with layers from 2 different logins. Connections stay at only 1 per user, which is half of 10.3. I haven't started on benchmarking this setup yet. But I like what I see so far. So your cautionary warning got me worried. What kind of circumstances do you mean? Maybe with any luck we do not work in those scenarios. Thanks. ... View more

Hi there again. I have been reviewing this and maybe I spoted the issue - still have to check... But I followed this article: https://blogs.esri.com/esri/arcgis/2012/03/08/an-alternative-to-overlaying-layer-tints-on-hillshades/ So there's some fiddling with gamma values and max/min values there and that might explain the different rendering. Going to try that when I get a bit of time. ... View more

Thanks all for the usefull suggestions! It got me exploring some options. First, when zoomed to raster resolution images look the same. Differences appear at zoom out, so that got me thinking on pyramids... just like Curtis suggested. The exported raster and the raster_with_function have the same settings: Resolution: 50m Pixel Depth: 8 Bit unsigned NoData Value: (empty) Pyramids: level 6, resampling: Nearest Neighbor Compression: None Display resample: None Stretch: None So I tried every combination of pyramids and display settings. The closest I got was building level 6, resampling: Cubic. It got me close but it's not the same thing still... bit darker and grainier. But close to the original. Really weird since the statistics of both rasters (dynamic with function vs exported) have the exact same statistics. Left: original, dynamic with raster function                      Right: exported, bicubic pyramids Differences are much less evident. So maybe arcmap is not showing the correct description of the dynamic with raster function and there's something done to the image that we don't know about. Anyway I'm going to accept these small differences. Thanks to all who chimed in. ... View more

Hi all. Yes this is potentially very good news. I am planning to test how now pgbouncer can be used with arcgis 10.4, desktop and server. This has the potential to drastically reduce the number of connections to more sane amounts. On the other hand, it may have significant impact on performance... thus the need to test. arcgis is very sensible to lan bandwidth. you can easilly see this if you put side-by-side a 100 Mbps and a 1Gbps machines with arcmap opening the same map. Don't even need to time it... ... View more

Hi there. You mention python health checks... can you share what you mean? thanks. ... View more

We had no issues if we stick with st_geometry, except for interoperability with other software (non-esri). pg_geometry (native postgis format) support is a bit sketchy sometimes. For instance, you cannot rename a feature class that uses the native spatial type. Which is really odd and annoying. The docs say there isn't a way to rename a spatial table in postgis so arcgis doesn't support it - this is a really awkward statement. We rename spatial tables all the time, so don't know what came over esri on this one... Also st_geometry does not support the same range of spatial aggregation operators of the native postgis format. But this is small use-case I guess. EDB Postgresql Plus is the same basic Postgresql plus some extra functionality but it's 100% compatible. So you can start with vanilla pgsql and then move to edb's if you want. I think it's an in-place replacement. Coming from Oracle, my biggest surprise was memory management. Postgresql relies on the OS cache management, so you will have to account for that when sizing/monitoring your server. Even if your connections take X GBs, there will have to be Y GBs free for the OS cache or your performance will suffer. Another thing I found different was the forthcoming of postgresql community. ... View more

On a side note, it seems many of us keep rasters in the gdb. Since arcgis started to use gdal, we have abandoned the raster-in-db and started using .vrt mosaics that point to the individual files, all kept in a file share. This is as good as a database, because we can update the individual files when needed. The only drawback is still the rebuilding of pyramids. We have not solved this. If a single image file is updated we have to rebuild the entire pyramids file (.ovr). In a GDB we can update a portion of the raster dataset and partially rebuild the pyramids. On the up side, we rarelly need to update files. Luckilly, what really happens is when get new imagery we usually get it in a new year/date and create a new mosaic, so rebuilding pyramids is really a non issue. The advantages are there is no rdbms maintenance nor server resources allocated to using raster data. Disk access and network bottlenecks still apply just as with raster-in-db. And backups are easier? ... View more

You've beat me about the lyr point... didn't see it until after I posted... ... View more